US 3866118 A
A microwave spectrometer includes means for sweeping an unknown sample and a standard sample with a microwave signal in a given bandwidth. The absorption spectrum of all the components, i.e., gases in the atmosphere, in the unknown sample which interact with the microwave signal and the absorption spectrum of the standard sample component, i.e., a pollutant gas which also interacts with the microwave signal are derived therefrom, the standard sample component comprising that component being analyzed in the unknown sample. The two derived absorption spectrum signals are cross-correlated for deriving a signal manifesting the amount of the component being analyzed, i.e., the pollutant, in the unknown sample.
Description (OCR text may contain errors)
United States Patent Ghosh et al. Feb. 11, 1975  MICROWAVE SPECTROMETER 3,351,762 V1373 ilieake 324/58 A X 3, 7 4 A  Inventors: Asoke Kumar Ghosh, D.D.O.; Harry 3 37 4 awor 32 /58 5 pdlerrefords both of Primary Examiner-Stanley T. Krawczewicz Que ana a Attorney, Agent, or FirmEdward J. Norton; William  Assignee: RCA Corporation, New York, NY. Squire  Filed: Nov. 8, 1973  ABSTRACT PP 413,809 A microwave spectrometer includes means for sweeping an unknown sample and a standard sample with a LS. Cl. A E E microwave signal in a given bandwidth. The absorp- 324/77 tion spectrum of all the components, i.e., gases in the  Int Cl G 27/04 atmosphere, in the unknown sample which interact  Field 58 A 77 with the microwave signal and the absorption spec- 23/254 E 255 trum of the standard sample component, i.e., a pollutant gas which also interacts with the microwave signal  References Cited are derived therefrom, the standard sample component comprising that component being analyzed in the UNITED STATES PATENTS unknown sample. The two derived absorption spec- 2,532,8l7 12/1950 Laffert), A trum signals are cross-correlated for deriving a signal 3,119,062 l/l964 Codd 324/58 A X manifesting the amount of the component being gg g z ig i lyzed, i.e., the pollutant, in the unknown sample. 3:522:527 8/1970 Williams, et al. 324/585 A 12 Claims, 2 Drawing Figures UNKNOWN 51 QORRELATOR F I I3 COMBINER Eto= rr x I H2 l7 DIFFERENCE Ai I a i I0; 6 I I SWEEP I cgmngm 52 2 OSCILLATOR E DIFFERENCE, A43 IA-BI I 26 2a 1 B 46 44 I E COMBINER I 22 DIFFERENCE MODULATOR 48 58 I 24 z I I F I STANDARD :5 060 HOMODYNE I 8 E l B 30 I A I l LOW I 60\ PASS I FILTER I I J T0 UTILIZATION MEANS MICROWAVE SPECTROMETER BACKGROUND OF THE INVENTION In microwave spectroscopy, the molecular composition of a gas is studied by determining the microwave frequencies in which absorption peaks occur. The basic apparatus which has been used for such study consists of a variable frequency microwave generator, a microwave crystal detector and a waveguide coupling the two together with the waveguide containing the gas to be analyzed. The absorption peak is then determined by scanning the frequency of the microwave generator across the range of interest and measuring the variations in microwave power at the detector. In this system extreme sensitivity of measurement is required. This sensitivity of measurement is compounded when the apparatus is used to analyze one gas contained within another gas.
In these systems, the measurement under discussion is that of a comparison of the microwave power received when absorption peak is present to that of the power received without any absorption peak. However, any variation in the ambient power level constitutes a noise background for the measurement. As a result, the prior art apparatus has approached the problem of microwave spectroscopy by attempting to reduce the noise for microwave spectroscopy measurements. In effect, noise reductions are attempts at increasing the sensitivity of the measuring instruments.
In these prior art microwave techniques only one frequency in a given spectrum of frequencies associated with a given sample component then being analyzed is detected at a given time. However, it is known that each pollutant may be defined bya microwave absorption spectrum of frequencies. Different pollutants or components are defined by different spectrums. It is therefore plain that a system which detects one frequency of a component is incomplete for identifying the corresponding spectrum of that component. Further, errors may be introduced in associating the one frequency with the corresponding spectrum since there is the possibility that there is one absorption frequency in the spectrum which is common to a number of components. As a result, discovery of a particular frequency of microwave energy which interacts with a component at best provides an ambiguous determination of the identification of that component in an environment including a plurality of components.
These techniques therefore require the search and analysis of a sample under examination at a number of separate, discrete frequencies and the production by trial and error as best as can be determined the spectrum for a given element in that sample. As a result, this procedure is complex and troublesome to implement. For a further discussion of the problems in the diagnosis of air pollution by microwave spectroscopy see the following article: Diagnosis of Air Pollution by Microwave Spectroscopy by William G'raff and I. H. Suffet, International Journal of Environmental and Analytical Chemistry, Vol. 1, pages 327-342, Gordon and Breach, Great Britain, I972.
SUMMARY OF THE INVENTION In accordance with the present invention, an apparatus is provided which includes means for deriving a first signal comprising solely a spectrum of frequencies of electromagnetic radiation within a certain bandwidth,
which spectrum interacts with a sample of matter. The sample includes a plurality of components including an unknown amount of a given component. Means are included for providing a second signal comprising solely the spectrum of frequencies of electromagnetic radiation within the certain bandwidth whichspectrum' interacts with a ,known amount of the given component. Correlating means are provided which are responsive to the first and second signals applied thereto for crosscorrelating the first and second signals, producing as an output signal therefrom a signal manifesting the amount of the given component in the sample matter.
IN THE DRAWINGS FIG. 1 illustrates a block diagrammatic illustration of an apparatus constructed in accordance with an embodiment of the present invention, and
FIG. 2 illustrates a second embodiment of the correlator of FIG. 1.
DETAILED DESCRIPTION In FIG. 1, sweep oscillator 10 provides a suitable microwave signal which sweeps across a predetermined microwave bandwidth. Preferably, this bandwidth is wide enough to emcompass all the major absorption lines of the given component. The oscillator 10 may be a conventional well known microwave generator.
The output of oscillator 10 is applied to two parallel branches l3 and 15. Branches l3 and 15 serve to provide at the respective outputs thereof microwave signals comprising solely the spectrum of frequencies at which absorption of the microwave energy occurs in the unknown and known samples. As to the known sample, the absorption spectrum is the characteristic spectrum for that sample. As to the unknown sample, the absorption spectrum may include the characteristic spectrum for each gas component contained therein.
The signals manifesting the absorption spectrum of the known and unknown samples are derived and correlated by cross-correlation techinques to be described below. As a result, the signal representing the absorption spectrum of interest is multiplied and integrated and at the same time isolated from the spectrum of the remaining components in the unknown sample as manifested by the microwave output signal of the unknown sample cell. The resultant cross-correlation signal is reduced to a DC level which can be readily calibrated to manifest the amount of unknown component in the sample being analyzed. Sensitivity and noise problems of prior;art techniques are overcome while providing improved sensitivity with reduced complexity.
In branch 13, the output of sweep oscillator 10 is applied by suitable waveguide means (not shown) through a variable resistance 12 to unknown sample cell 14. Cell 14 is a conventional unit which holds a sample of gas such as the atmosphere or the like to be analyzed. The output of cell 14 is applied to a combining device 16 through a variable resistance 18 by way of suitable waveguide apparatus (not shown). The output of oscillator 10 is also applied to device 16 through variable phase shifter 17 coupled across cell 14 and resistance 18.
Device 16 may be a conventional so-called waveguide magic T device which has two inputs and two outputs. At one output a signal is provided which is the sum of the two input signals applied thereto, while at the other output a signal is provided which is the differ- 3 ence of the two input signals applied thereto. The device 16 output signal E (signal A) appearing on line 20 is the difference output, representingthe difference between the device 16 input'signals E and E The signal E, is the signal produced by cell l4 and resistance 18. The signal E} is the output signal of phase shifter 17. l I
smaller than the output signal corresponding to the In branch 15, the output of sweep oscillator is applied to a standard sample cell 22 by way of Suitable waveguide apparatus (not shown) through a variable phase shifter 24. The output of phase shifter 24 is also applied to variable phase shifter 26. The outputsignal E of phase shifter 26 is applied as an input to a socalled waveguide magic T device -28 as one input and the output of cell 22 is applied through variable resistance 30 as a second inputsignal E to device 28. The output of device 28 along line 32 is signal B also;designated E Signal E represents the difference between the two input signals 13,; and E applied as separate inputs to magic T device 28. The signal E (A)" on lead and signal E (B) on lead 32 are applied to ,crosscorrelation correlator 34. Correlator 34 by crosscorrelation provides at the output lead 36 thereof'a sig-.
nal which represents the analyzedamount of the prede cornponent in the gas sample to; be measuredgTo do" termined gas component present in the sample in unknown cell .14.
Correlator 34 includes modular 46 and combiner 40 Modulator 46 modulates signal B at a suitable frequency provided by os cillator48'. Preferably this fre-" quency is mHz. The modulated output signal B of a modulator 46 is applied as one of two inputsto com biner 40. The other signal applied as an input to combiner 40 is signal A. The combiner 40 output signal on lead 42 is (A+B), while the signal on lead 44 is A-B). It should thus be appreciated that combiner 40 maybe a conventional so-called magic T device whichl takes the sum and difference between the two input signals A and B; l l 4 Output lead 42 is coupled as an input to squar er 50 while output lead 44 is coupled as an input to squarer 52. Squarers 50 and 52 may be of thediode detector two inputs. The output ofcombiner 54 is a signal con- 6 type. The outputs of squarers 50 and 52 are applied as inputs to combiner 54, Combiner 54 is another so- .Called magic T which takes the difference between the e l taining a DC component and a component that is twice the frequency of oscillator l0.-Both of these compo nents are modulated at the modulating frequency 30 mHz because of the modulation of signal B. The modulation is detected and applied to IF amplifier 56, the output of which, in turn, is applied as an input to homodyne circuit 58. Circuit 58 receives as an input a signal from oscillator 48. I-lomodyne circuit 58 produces in a well known manner at the output thereof a DC signal representing the amount of unknown component in cell 14. The DC signal is then applied to low pass filter 60 device 16,
The variable resistance 1 2 isprovided to equalize the amplitude of the sweep signal applied to branches l3 i and 15..The amplitudeofthe signal appliedto branches l3 and 15 .shouldbe the same within agiven amount 1' which amount results iii a signal on lead36 thatis smallest amount of the component in the gassample to be measured. To do this is within the skill of those skilled in this art.
Phase shifter 24 serves tosetthe phase of the microwave signal applied to branches l3 and 15 insubstantially identical phase. That is, phase shifter 24 is pread-* justed so that the signal B' on branch 15 output lead 32 I is substantially identical to the phase of signal A on branch 13 output lead20. It will be appreciated by. those skilled inwthis art lthat the phase in the two branches maybe off slightlywith respect toeach other.
The phase of the "signals appliedtto brariches 13 1and;:15
l should be the same withiii a given amount which amountlresultsin a signal on correlatoi' output leady36 that. is smaller in amplitude than the output signal on 6 lead 36 Correspondingrft o. thew'smallest amount of the this is also within the skillof those skilled in this art. .As
a resultl the sweep signals appliedtothe respectivelegs of branches 13 and l5;are substantially .the stime phase and amplitude. In: addition, the phase and amplitude of the signal processed in eachleg I.of branch 13 a" need also be substantiallythe same. This criteria also nalE substantially identical to the amplitudeof signal Ers. I a V1 In branch 13, in a 'milar manner, variable resistance 18 andphaseshifte r. 117 are preadjusted to set the respective amplitude'and phese of the respective output signals E; andE on the legs of branch llssubstantially identical toeach other il he phase shif ger. 17 output sigcombined with thevariable renal E in branch 13; 6 sista hce IS outputsignaI E; by combiner devi ce l6 lni noted above.
I The physical length at eachside ofbrancli 13 including cell 14, resistance 18on one side and variable phase shifter 17 on the other sideare substantially the same. The leg of branch l3 including phase shifter 17 is made substantiallyv the sameelectronic length as the other leg 5 by phase shifter 17. In the same manner, the lengths of whose output 36 is applied tosuitable utilization means each leg of branch 15 are made substantially thesame. It will thus be evidentthatsignal E in branch 13 and signal E in branch 15 comprise the respective signals provided by sweep oscillator 10 altered by the absorption spectrum of the compbnents in respective cells 14 and 22.As is well knowmthe absorption of microwave energy by a sample reduces the power. level thereof at each of the frequencies inthe absorption spectrum. However, iniaccordance with therpresen-t invention, the absorption spectrum isprovided in terms of increased power level by subtracting signal E from signal E and signal E from signal E Further, the sensitivity of the spectrometer of the presentinvention is also enhanced by: this subtracting process wherein there is thus proapplies to the phase aii d amplitude of the signals in] Teach leg of branch ,15. s l s Inbranch 15 phase shifter 26 adjusts the phase of output signal E, substantially identical in phase with output signal E produced. by serially cqnnected stan dard sample cell 2 2 and variable resistance .30. Variable resistance 30serv es to adjustthe amplitude of sigsignalE is subtractedllfrom signal Et, as t vided signals E and E which include the absorption spectrum of the unknown sample. For each frequency at which absorption occurs in cells 14 and 22, the difference between the signal levels of signals E and E and signals E and E, respectively, provide positive excursions. These positive excursions define the spectrum of the unknown components in cell 14 and the known component in cell 22.
A standard sample of gas stored in cell 22 is a sample of a known amount of one component of the gas to be examined in the unknown sample cell 14. Thus, while the cell 14 may include a number of components each having a correponding different absorption spectrum, the sample contained in cell 22 is a single component having a single characteristic absorption spectrum. It will thus be appreciated that the signal E, is a signal having a characteristic spectrum for a single component of the gas contained within the standard sample cell 22. This characteristic spectrum may or may not be present in the spectrum of the unknown sample signal- E It should be appreciated that the signals E and E are approximately zero in value except at those frequencies at which absorption occurs.
It can be shown mathematically that a crosscorrelation term is produced by cross-correlation of signals E and E This term is represented by a signal having a parameter which is proportional to the product of the amount of the known component or gas in cell 22 and the unknown amount of the same gas or element in sample cell 14.
With no gas in cells 14 or 22, signals E E E and E, are adjusted to be substantially the same in amplitude and phase in the manner described above. It can be shown in this case that the following relationships are present:
E1 ERXe -a L uwz XL) s ras SLzEnsu a 1.)
where 0: and a, contain both the absorption coefficient of the gas and the concentration of the gas in the cell, and e M has been approximated by l "01L, and L is the length of the line in each side of the branches l3 and in centimeters.
Combining at devices 16 and 28:
10 RX .r)/ V7: CY L/ V5 E30 ns s)/ V7: ns S VT Combining at device 40:
E, Vz(E a L -E a,L) cos wt E /2 (E a L-l-E a L) cos wt where the time variation at the microwave frequency has been included. Squaring and substituting (l+cos2wl)/2 cos wt EAZ /8 +cOS2wt) (Enx X2 L2 l'E (15 14 2 2ERSERX s 1 E i /s (1+cos2 i wt) (E a L +E afL -l-2E E a, a L
E E A (l+cos2wt) (E E ,a,a,L neglected terms.
As seen in equation 9, the cross-correlation term shown is isolated from other terms (neglected) containing only absorption in the standard cell or only absorption in the unknown cell. The cross-correlation term appears as a DC term and also as an AC term at twice the carrier frequency, that is, twice the frequency of the resonance, or, if a heterodyne receiver is used, at twice the IF frequency.
As set forth in equation 9 the signal on lead 36 comprises solely the correlation term manifesting the signal whose amplitude is proportional to the amount of the unknown gas in sample cell 14. The amplitude of the DC signal at output 36 can be calibrated in a conventional manner to provide a suitable indication of the amount of unknown gas in cell 14.
Since the signal on output lead 32 of branch 15 having a given phase and bandwidth representsa known sample, such a signal can be provided by a suitable prerecorded apparatus such as a tape recording or other storage means.
The sample of gas to be analyzed placed in cell 14 is at reduced pressure. In providing the gas to be analyzed, cell 14 is evacuated and the sample of the atmosphere being tested placed in the cell 14. This is accomplished by conventional well known pumps and valves.
In the embodiment of FIG. 1, two squarers 50 and 52 are utilized. These squarers need to be precisely matched. The extent to which these devices need be matched can be determined in a conventional manner. However, to avoid the matching problem presented in the embodiment described above, an alternate arrangement can be provided.
In the alternative arrangement, FIG. 2, a single square law detector is provided in correlator 34 which therefore avoides the matching problem. In particular, this latter arrangement includes a bi-phase modulator 72 driven by an oscillator 74. The modulator 72 receives signal A at an input 76 thereto. The biphase modulator 72 changes the phase of the signal A between 0 and 180 for each cycle of the modulator driver frequency. Thus, the phase is 0 for one-half cycle and l for the other half of each cycle of the driver frequency.
The bi-phase modulated signal is combined in a magic T device 78 with signal B on input 77. The magic T output signal appearing at output 80 contains the sum signal (A+B') when the modulator provides 0 phase shift and the difference signal (AB') when the modulator 72 provides phase shift. The magic T 78 output signal is applied to square law detector 70.
The sum and difference signals are time shared at the frequency of the modulator driver oscillator 74.
The sum and difference signals differ in magnitude resulting in a ripple on the output 82 of the square law detector 70 at the frequency of the modulator driver. l
This ripple contains only the cross-modulation term. In turn, the ripple is amplified by an amplifier 84 tuned to e (EH2) ct,,0 L
where r nx ns- Thus, the alternate arrangement of FIG. Zprovides a cross-correlation term, equation 10, while using a single receiving channel and providing fully coherent detect'ion.
What is claimed is:
l. in combination:
first sample-holding means for holding a first sample of matter having a known amount of a given component,
second sample-holding means for holding a second sample of matter having a plurality of components including an unknown amount of said given component, l means for applying microwave energy to each said sample-holding means, said energy being absorbed by at least said given component in a spectrum of frequencies, a different absorption spectrun being exhibited by each different component absorbing said energy, each said sample-holding means prothe modulating frequency. The amplified ripple is then means responsive to said.microwave energy applied thereto coupled across said second sample-holding means for setting said applied microwave energy in substantially identical phase with said second sampleholding means output signal, and means for subtracting said second sample-holding means output signal and said set microwave energy across said second sampleholding means from each other to produce said second signal.
4. The combination of claim 3 wherein one of signal producing means includes phase setting means for setting substantially identical the phase of said microwave energy applied to said first and second signal producing means.
5. In combination:
means for deriving a first signal comprising solely a spectrum of frequencies of microwave energy which spectrum is formed by absorption of certain said energy by a sample of matter, said sample in cluding a plurality of components including an unknown amount of a given component,
means for deriving asecond signal comprising solely a spectrum of frequencies of microwave energy 6. The combination of claim 5 wherein said first signal deriving means includes a first sample-holding cell ducing a respective output signal corresponding to said microwave energy altered by said absorption spectrum,
means responsive to said microwave energy and said output signals applied asrespective inputs thereto for producing a first signal manifesting solely the spectrum of microwave energy absorbed by said first sample and a second signal manifesting solely the spectrum of microwave energy absorbed by said second sample, and
correlating means responsive to said first and second signals applied as inputs thereto for crosscorrelating said first and second signals to thereby produce a signal manifesting the amount of said unknown component in said second sample.
2. The combination of claim 1 wherein said first signal producing means includes first phase setting means responsive to said microwave energy applied thereto coupled across said first sample-holding means for setting said applied microwave energy in substantially identical phase as said first sample-holdingmeans output signal, and means for subtracting said first sample holding means output signal and said set microwave ent ergy from each other to produce said first signal.
3. The combination of claim 2 wherein said second signal producing means includes second phase setting and first phase adjusting means coupled across said cell, said first sample-holding cell and said first adjusting means including input means for receiving said microwave energy, said first cell and first adjusting means each having an output signal comprising microwave energy having substantially identical phase, and combinsaid second cell, said second cell and second adjusting means including inputmeansfor receiving said microwave energy, said second cell and second adjusting means each having an output signal comprising microwave energy having substantially identical phase, and
combining means responsive to said cell and adjusting ing means includes means responsive to said first and second signals applied as inputs thereto for taking the sum and difference between said first and second sig nals to provide as output signals therefromthird and fourth signalscorrespon ding to said sum and difference, squaring means for squaring said' third and fourth signals applied thereto, and means responsive to said squared thiid and fourth. s ignals for deriving a fifthsignal manifestingthe product of said first and second sigcell and second phase adjusting means coupled across U 9. The combination of claim.8 wherein said means for taking the sum and difference includes means for modulating one of said first and second signals.
10. The combination of claim wherein said first and second signal deriving means each includes a separate, different sample-holding means coupled in parallel with corresponding phase adjustingmeans, and said deriving means further includes means for applying microwave energy to said sample-holding means and said phase adjusting means.
11. The combination of claim 10 wherein said applying means includes a resistance serially coupled to one of said first and second deriving means and phase adholding means.
UNITED STATES PATENT OFFICE CERTIFICATE OF CORRECTION Patent No. 3, 66,118 Dated February 1'] 1975 Inven r( Asoke Kumar Ghosh and Harrv John Moodv It is certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:
L) (2r should read I 1 =5 L 1 c L U (1 L (2) C01. 6, lines l-3 H a 2 A 2, 2 2 2 2 LA 1/8 (1 cos/.wt) (L o0 L L a L 9 A -ZIZ E u Q L) 7 should read v.Z 2 I m l/s (1+cos2wt) (Eg af +L; ca L -ZL L Q QL L (7) Col 6, lines 6 i- 7 7 2 2 2 I3; 1/8 (1+cos2 1 wt) (E (1 L +1= c: 2 +2 H o. a L 2 RS 5 L 'RSRX S 1 should read 2 L +F 031. a a L (s Z I w2 Er 1/5 (1+cos2wt) [h u 5 x h S Col. 8., line 41, after "signals of should read for Signed and sealed this 27th day of May 1975.
C. MARSHALL DANN RUTH C. MASON Commissioner 'of Patents Attesting Officer and Trademarks FORM Po-105O (10 69) uscoMM-Dc 60375-3369 U 5. GOVERNMENT PRINTING OFFICE: a 69- 930